Classification of breast lesions presenting as mass and non-mass lesions

We aim to develop a CAD system for robust and reliable differential diagnosis of breast lesions, in particular non-mass lesions. A necessary prerequisite for the development of a successful CAD system is the selection of the best subset of lesion descriptors. But an important methodological concern is whether the selected features are influenced by the model employed rather than by the underlying characteristic distribution of descriptors for positive and negative cases. Another interesting question is how a particular classifier exploits the relationships between descriptors to increase the accuracy of the classification. In this work we set to: (1) Characterize kinetic, morphological and textural features among mass and non-mass lesions; (2) Examine feature spaces and compare selection of subset of features based on similarity of feature importance across feature rankings; (3) Compare two classifier performances namely binary Support Vector Machines (SVM) and Random Forest (RF) for the task of differentiating between positive and negative cases when using binary classification for mass and non-mass lesions separately or when employing a multi-class classification. Breast MRI datasets consists of 243 (173 mass and 70 non-mass) lesions. Results show that RF variable importance used with RF-binary based classification optimized for mass and non-mass lesions separately offers the best classification accuracy.